Incorporation of N-Heterocyclic Carbenes and Their Precursors into Metal-Organic Frameworks The University of Adelaide School of Physical Sciences Department of Chemistry Submitted in fulfilment of the degree Master of Philosophy (Chemical Science) Presented by Patrick Capon B. Sc. (Advanced) Supervisors: Associate Professor Christopher Sumby and Associate Professor Christian Doonan October 2016
119
Embed
Incorporation of N-Heterocyclic Carbenes and Their ...Incorporation of N-Heterocyclic Carbenes and Their Precursors into Metal-Organic Frameworks The University of Adelaide School
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Incorporation of N-Heterocyclic Carbenes
and Their Precursors into Metal-Organic
Frameworks
The University of Adelaide School of Physical Sciences
Department of Chemistry
Submitted in fulfilment of the degree
Master of Philosophy (Chemical Science)
Presented by
Patrick Capon B. Sc. (Advanced)
Supervisors: Associate Professor Christopher Sumby and Associate Professor Christian Doonan
October 2016
i
Declaration
I certify that this work contains no material which has been accepted for the award of any other degree
or diploma in my name, in any university or other tertiary institution and, to the best of my knowledge
and belief, contains no material previously published or written by another person, except where due
reference has been made in the text. In addition, I certify that no part of this work will, in the future,
be used in a submission in my name, for any other degree or diploma in any university or other tertiary
institution without the prior approval of the University of Adelaide and where applicable, any partner
institution responsible for the joint-award of this degree.
I give consent to this copy of my thesis, when deposited in the University Library, being made available
for loan and photocopying, subject to the provisions of the Copyright Act 1968.
I also give permission for the digital version of my thesis to be made available on the web, via the
University’s digital research repository, the Library Search and also through web search engines, unless
permission has been granted by the University to restrict access for a period of time.
Patrick Capon, 31st of October, 2016
ii
iii
Abstract Metal-organic Frameworks (MOFs) are a class of porous materials with excellent potential for
application in catalysis, gas storage or molecular separations. MOFs are synthesised by combination of
an organic linker unit and metal node precursor to yield an overall network structure that typically
extends in two or three dimensions. Often the network contains void space, which is the origin of the
large surface areas and high porosity observed for many MOFs.
N-Heterocyclic Carbenes (NHCs) were originally applied as ligands for metal complexes, and
are commonly used as supporting ligands for organometallic catalysis. For example, an NHC is
incorporated into Grubbs’ second generation catalyst. NHC precursors have been incorporated into
MOFs, leading to properties that make them applicable to catalysis and gas sorption. Metalation of the
NHC precursor to yield a MOF bound NHC-metal complex provides an opportunity to further enhance
a MOF’s capacity for gas sorption or to provide a site for catalysis to be performed.
The main aim of this thesis was to incorporate NHCs into MOFs to yield materials with
applications in catalysis (via NHC metalation) or gas sorption. In Chapter 2 five new azolium or NHC
containing 1-D MOFs are presented (1-Cu, 2, 3, 4, and 5), with 1-Cu and 2 showing strong enthalpy of
adsorption values for H2 gas. An NHC-Cu(I) complex was generated concomitantly with MOF synthesis
to yield 1-Cu, however this metal site was not viable for catalysis due to the low porosity of 1-Cu.
These studies were extended to the highly stable DUT-5 MOF in Chapter 3 with two new DUT-5
analogues and two new mixed linker DUT-5 analogues generated. All four materials showed improved
porosity compared to the MOFs in Chapter 2. In Chapter 4, the DUT-5 analogues were investigated for
CO2 and CH4 gas sorption properties at high pressure in order to provide increased industrial relevance.
Furthermore, ionic liquids (ILs) were included within the DUT-5 analogues in an attempt to improve
CO2 uptake. However, excessive loading of the IL resulted in a loss of MOF porosity and minimal uptake
of CO2 or CH4. Finally, early metalation outcomes of the NHC precursor containing DUT-5 analogues
are discussed in Chapter 4, with an aim toward NHC-metal based catalysis in further experiments.
iv
Acknowledgments
Firstly I wish to thanks my supervisors Chris Sumby and Christian Doonan, who have supported me a
lot over the past two years. A special mention is due to Chris, who has put in countless hours helping
me with anything ranging from experimental problems, crystallography, or general advice, let alone
the incredible amount of renditions of this work he has read. I greatly appreciate all the help Chris
has given me, and I certainly wouldn’t have produced anything of substance without his assistance.
Secondly thanks are due to the entire Sumby-Doonan group over the last two years, who have
shared some or all of this journey with me. To Alex Burgun, thank you for showing me the ropes in
the lab to begin with, and I am very glad for the strong friendship that we have developed (and I’m
still winning the Pokémon battle count – just!). A similar thanks goes to Cam Coghlan and Jesse Teo,
who spent a lot of time early on teaching me all manner of experimental techniques – I am indebted
to both of you. To Natasha Zaitseva and Michael Huxley, thank you for your company and advice
within the lab, it wouldn’t have been the same without you. To Kate Flint, thank you for making this
last year much more bearable and a whole lot of fun. I’ve thoroughly enjoyed working with you and I
doubt there could ever be as clean a fume cupboard again without us around. To Harley Betts,
Natasha Maddigan, and Oliver Linder-Patton, thanks for being excellent friends and lab mates – I
couldn’t have made it through without your support.
Thanks are also due to my friends and family who provided me with refreshing conversation and
helped me to keep everything in perspective. Specifically to my parents Deb and David – who have
supported me immeasurably for as long as I can remember – thank you for everything.
To my partner Aimee, I most certainly couldn’t have made it through the last two years without you
by my side, thank you for your dedication and unwavering support.
Table of Contents Declaration .............................................................................................................................................. i
Abstract ................................................................................................................................................. iii
Acknowledgments ................................................................................................................................. iv
Abbreviations ......................................................................................................................................... v
benzenedicarboxylate) family of MOFs, with 1 and 2 exhibiting higher Qst values than the Zn and Mn
frameworks, but not reaching that of the Ni framework.82 Furthermore, the Qst values are comparable
with those obtained for anionic zeolite-like MOFs.83 This suggests that micropores lined with a charged
imidazolium group are a competitive option in the field of H2 adsorption. Interestingly, while the Qst
values for 1 and 2 are initially higher than 1-Cu, the metalated framework retains a larger enthalpy of
adsorption at increased H2 loading. These high Qst values are consistent with narrow pore widths for
1-Cu (6.43 Å) and 1 and 2 (6.79 Å),27 as determined by Non-linear density functional theory pore size
calculations (Fig. 2.14).
Fig. 2.13: a) H2 isotherms at 77 K for 1 (red), 1-Cu (blue), and 2 (green). b) Isosteric heat of adsorption for H2 for 1 (red), 1-Cu
(blue), and 2 (green).
33
Fig. 2.14: Pore size distribution calculated by Non-linear density functional theory for 1 (red), 1-Cu (blue), and 2 (green).
Table 2.1: Low coverage Qst values obtained from proprietary software on the Micromeritics 3-Flex Analyser.
MOF Low coverage Qst value (kJ/mol)
1 -9.189
1-Cu -8.999
2 -9.093
2.2.2 Toward azolium and NHC containing MOFs with increased porosity
Given the chemical versatility of H3L1 to produce four new MOFs from Cu(II), Co(II), and Mg(II) nitrate
salts and Mn(II) chloride, H3L2 was also screened against a series of metal salts to investigate its
coordination chemistry. The para positioning of the carboxylate groups in H3L2 as opposed to the meta
positioning in H3L1 was predicted to yield structures with a more open network. Unfortunately H3L2
did not form any new MOFs from this widespread testing other than with Zn(NO3)2, of which two MOFs
have already been reported in the literature. The solubility of H3L2 in DMF is significantly lower than
H3L1, with H3L2 precipitating from solution at temperatures below ~80°C. On the other hand H3L1
readily dissolved in DMF at room temperature, which likely accounts for the increased versatility of
MOF formation, under common MOF synthesis conditions, for H3L1 as compared to H3L2.
The first Zn(II) based MOF using H3L2 was reported by Sen et al.65e and incorporates a
purported Zn8O cluster and exhibited high proton conductivity. The second was reported by Burgun et
al.6 and was able to be metalated with Cu(I) in order to catalyse the hydroboration of CO2. Given these
34
unique results, a further investigation of Zn(II) based MOFs from H3L2 was undertaken, with an initial
aim toward in situ metalation of the NHC precursor with Ag or Au. Unfortunately no metalation was
achieved, however two separate MOFs were obtained from synthesis conditions which differed only
in the volume of DMF used (Table 2.2). MOF 6 was originally synthesised in the presence of AgNO3 in
attempt to metalate the NHC precursor with Ag, however it was confirmed to form without AgNO3
present. MOF 7 was synthesised in a greater volume of DMF and crystallised in the P-3 space group as
opposed to R-3 for 6. Preliminary structural data for both 6 and 7 in comparison to the structure of
Sen et al.65e is discussed in more detail below.
Table 2.2: Synthetic conditions for MOFs 6 and 7. Note that the metal salt used was Zn(NO3)2.6H2O, and all syntheses were
conducted at 100°C for 24 h.
MOF code H3L2:Zn molar
ratio
Linker concentration
in DMF Additive Result
6@Ag 1:8 5 mg/mL AgNO3 1 equiv. Rectangular crystals in
precipitate
6 1:8 5 mg/mL None Rectangular crystals
7 1:8 2 mg/mL None Hexagonal crystals
The original Zn(II) MOF reported by Sen et al.65e consists of 2-D sheets (Fig. 2.15) which are
then 3-fold interpenetrated to form a 3-D porous material (Fig. 2.16, Fig. 2.17). The MOF crystallises in
the R-3 space group, and forms a 6,3 connected net. The asymmetric unit that yields the 2-D sheets
consists of two H3L2 molecules bound to a tetrahedral Zn centre, with bond distances of 1.95 Å. A DMF
molecule is also coordinated to the Zn centre with a bond distance of 2.00 Å. Two further Zn atoms are
included in the asymmetric unit to form a Zn8O node. This node is then bound to six H3L2 molecules in
a spiral formation, with a ‘radar dish’ motif attached to each linker arm to form a 48 membered
metallocycle. 6 is isostructural to the Zn(II) MOF synthesised by Sen et al.,65e with the exception of the
metal node as discussed below (Fig. 2.18).
35
Fig. 2.15: Tilted view of the 2-D sheet present in the Zn(II) MOF synthesised by Sen et al. The 2-D sheet propagates to the left
and right in this view. Zinc, light blue; carbon, grey; nitrogen, dark blue; hydrogen, white.
Fig. 2.16: Three interpenetrated nets of Sen’s Zn(II) MOF represented in red, green, and purple. Hydrogens are omitted for
clarity.
36
Fig. 2.17: Space filling view down the a axis of the Zn(II) MOF synthesised by Sen et al., where the positioning of the
imidazolium group on the pore channel walls is highlighted. Figure produced by Sen et al.65e
Fig. 2.18: View of the Zn metal node and six nearest ‘radar dish’ motifs in Sen’s MOF and 6 both a) top down, and b) side on.
Zinc, light blue; carbon, grey; nitrogen, dark blue; hydrogen, white.
The SCXRD data obtained here for MOF 6 disagrees with the assignment of the Zn8O node by
Sen et al.,65e instead suggesting a disordered Zn4O metal node (Fig. 2.19). This type of node has been
observed in literature previously by both the Lin84 and Zhou85 groups. This is evident by comparing the
thermal ellipsoids obtained when the structure is assigned with a Zn8O node or two disordered Zn4O
nodes (Fig. 2.20a and b, respectively). The thermal ellipsoids for the oxygen atoms in the Zn8O model
are elongated and significantly larger than those of the Zn atoms (Fig. 2.20). This disproportionality of
ellipsoid size for the Zn centres suggests that too little electron density has been modelled in this
region. On the other hand, all ellipsoids in the Zn4O model are similar in overall disposition and the
oxygen ellipsoids are not elongated, which indicates an improved modelling of the electron density.
Hence here the disordered Zn4O node is suggested as the correct assignment of the electron density
37
obtained from the diffraction data. While the node composition does not affect the overall packing of
MOF 6, it does affect the molecular formula and density of the material. Thus, properties such as
proton conductivity and porosity may be affected and should be re-evaluated.
Fig. 2.19: The two potential nodes for MOF 6; a) the Zn8O node proposed by Sen et al.65e and b) the two disordered Zn4O nodes
proposed here. Both primary nodes are shown in yellow, and the alternate node in b) is shown in red and blue. Zinc, light blue
or yellow; oxygen, red or yellow (in b only); carbon, grey.
Fig. 2.20: Thermal ellipsoids for a) the Zn8O model, and b) the disordered Zn4O model indicating the improved electron
density fit obtained for the Zn4O model, as discussed above.
Another intriguing difference between 6 and the structure published by Sen et al. lies in
collection of PXRD data. Sen et al. showed retention of crystallinity up to 523K by PXRD for their Zn8O
material, however here 6 was unstable to solvent loss, resulting in a loss of crystallinity when the PXRD
pattern was collected (Fig. 2.21). While this is preliminary data, it is worthy of note as this may then
affect the properties of the bulk material. The synthesis conditions here were identical to Sen et al.,
but evidently there is some point of difference between the two materials that is yet to be discerned.
38
Fig. 2.21: PXRD data for Sen et al., simulated (black), 6 simulated (blue), and 6 desolvated (purple), showing the loss in
crystallinity upon desolvation of 6.
MOF 7 was originally synthesised by Lewis (University of Adelaide Honours thesis, 2011) but
not published due to the unreliable synthetic procedure used at the time. MOF 7 is similar to 6, with
the key difference being the quantity of DMF used in synthesis, while all other synthetic components
are held constant. DMF was observed to bind to the tetrahedral Zn centres of the ‘radar dish’ motif in
7 and all linkers are bound in a monodentate fashion. In comparison, 6 has one carboxylate on each
linker bound in a bidentate fashion, and the other remains monodentate to form the three tetrahedral
Zn centres. While 7 also forms interpenetrated 2-D sheets, there are only three metallocycles bound
to each node as opposed to six rings in 6 (Fig. 2.22). This causes the interpenetration to now be 6-fold
to form a 3-D porous material (Fig. 2.23). The increase in DMF appears to inhibit formation of the Zn4O
node, and thus causes this structural difference. However, the exact nature of the compound bound
to the tetrahedral Zn could not be determined with the diffraction data obtained, but appears to be
an oxygen donor (Fig. 2.24). It is predicted to be disordered DMF or water based upon the reaction
conditions. One nitrate anion is observed within the pore space of 7, however an additional nitrate is
required to achieve charge balance that is not observed crystallography. Despite these crystallographic
issues, the connectivity and overall MOF structure are well established and are a reasonable model of
the electron density map.
39
Fig. 2.22: 2-D sheet of 7, showing only three metallocycles originating from each Zn centre. The sheet propagates to the left
and right in this view.
Fig. 2.23: Six interpenetrated nets of 7 shown in red, green, light blue, purple, yellow, and brown.
Fig. 2.24: View of the zinc node in 7, where disordered solvent (top O donor on tetrahedral Zn) binds instead of three linkers,
thus inhibiting formation of the Zn4O node as observed in 6. Note that the MOF continues to propagate in the radar dish
motif at the end of each linker, but this is not shown here. Zinc, light blue; carbon, grey; nitrogen, dark blue; hydrogen,
white.
40
Removal of the Zn4O node in 6 to form 7 results in a change of space group from R-3 to P-3,
and thus the simulated PXRD pattern also differs (Fig. 2.25). As observed for 6, 7 was unstable to
solvent loss and therefore only the simulated PXRD patterns are compared here.
Fig. 2.25: Simulated PXRD patterns for 6 (blue) and 7 (red) which demonstrate the change in space group from R-3 to P-3.
2.2.3 Imidazolium Containing Linker Substitution into a Porous MOF As discussed in the introduction, substitution of an azolium containing linker into a parent MOF is a
common alternative to direct MOF synthesis. This can often lead to ML-MOFs with new topologies that
are otherwise inaccessible by the conventional synthetic route. Here a linker synthesised by Kong et
al.19c (H4L3) was identified as a candidate for substitution into the well-known MnMOF1 framework,
which has shown considerable chemical versatility for post-synthetic modification and ligand
substitution.22, 46 H4L3 was chosen as it is of a similar length and design to the native H2L4 linker used
in MnMOF1. Furthermore, the structure of MnMOF1 consists of 2-D sheets which are bound to form
a 3-D material by H2L4 moieties which bind to Mn through the carboxylate donors alone, whereas the
linkers within the 2-D sheets bind with both carboxylate and N-donor groups. Thus, H4L3 was proposed
to be capable of replacing these H2L4 moieties which bind only through carboxylate groups.
A series of one pot syntheses and solvent assisted linker exchange (SALE) reactions were
attempted, however none of these resulted in a crystalline product. From the reactions attempted, it
was evident that H4L3 easily precipitates out of solution in DMF, likely due to the two charged
imidazolium groups present on the molecule. While H4L3 is readily soluble in water, any SALE reactions
involving water resulted in the destruction of any MnMOF1 crystals and H4L3 remained in solution.
Interestingly, the one pot syntheses utilising water still resulted in a H4L3 precipitate (determined by
41
1H NMR spectroscopy and PXRD) but formation of MnMOF1 was only observed when the MnMOF1
linker was in large excess. This suggests that H4L3 inhibits growth of pure MnMOF1.
H4L3 was chosen as a candidate to replace the carboxylate donor H2L4 moieties in MnMOF1 as
discussed above. However, given the lack of successful SALE or one-pot syntheses, H4L3 appears to not
be an ideal substitution candidate. This is likely due to the charge balance required in MnMOF1. When
deprotonated, L4 has a 2- charge due to the carboxylate groups, while L3 is neutral as the negative
carboxylates are balanced by the two positively charged imidazolium groups. Thus a neutral linker is
attempting to replace an anionic linker, which will result in an unfavourable charged environment.
Furthermore, close examination of the linker shape when in reported MOFs suggests another source
of incompatibility (Fig. 2.26). When crystallised as part of the Cu(II) MOF synthesised by Kong et al.,19c
L3 twists about the methylene bridge to yield a reverse S like shape. On the other hand, L4 also twists
around the methylene bridge in MnMOF1 (solvated with DMF), but this causes both carboxylate
groups to exist on the same side as each other. This suggests that L3 may not be preferentially able to
obtain the shape required to substitute for L4 in MnMOF1.
Fig. 2.26: Crystallised structures of L3 and L4 from their native MOFs (extended structure not shown). L3 is twisted into a
reverse S type shape, while L4 twists such that both carboxylates are on the same side as each other. Carbon, grey; nitrogen,
blue; oxygen, red; hydrogen, white.
42
Given the issues discussed, no further attempts at substitution reactions of H4L3 into MnMOF1
were made. Instead this section acted as a preliminary investigation into SALE that was used as a
stepping stone toward the substitution reactions that are described in Chapter 3.
2.3 Experimental
2.3.1 General Methods
Unless otherwise stated, all chemicals were obtained from Sigma-Aldrich and used as received. 1,3-
DUT-5 is a viable MOF system that allows for the NMR digestion and pore space issues
discussed above to be circumvented. Samples of DUT-5 can be digested in a D3PO4/d6-DMSO solution,
thus providing the ability to monitor bpdc-Im or bpdc-Me incorporation in ML-MOFs. Furthermore,
the pore apertures of DUT-5 are 11.1 x 11.1 Å wide,15 and are expected to provide increased space
around the imidazolium moieties of bpdc-Im, thus potentially allowing for improved access for
metalation or increased gas adsorption volume. With this in mind, the major focus of this research
shifted to the DUT-5 system.
3.2.4 DUT-5 based systems DUT-5 was synthesised with relative ease using the method of Senkovska et al. and did not require
large quantities of modulator as for the UiO-67 system.15 Using the published conditions as a starting
point, two new DUT-5 analogues, DUT-5Me and DUT-5Im, were synthesised from linkers bpdc-Me and
59
bpdc-Im as detailed in the experimental section (3.5.3). PXRD data indicated a successful synthesis by
good agreement with the simulated pattern for DUT-5 with only minor differences in relative intensity
(Fig. 3.7). The presence of both bpdc-Me and bpdc-Im was further supported by NMR digests (as
detailed in the section 1.5.1) of DUT5-Me and DUT-5Im, which contained diagnostic signals for both
linker types respectively (see Fig. 3.8 for key signals, full NMR spectra in the Appendix, Fig. A2.1 and
Fig. A2.2).
Fig. 3.7: PXRD for DUT-5 simulated (black), DUT-5 experimental (red), DUT-5Me (green), and DUT-5Im (blue). Note the
background is not subtracted here as this provides a better view of the low intensity peaks.
Fig. 3.8: Key 1H NMR signals for bpdc-Im and bpdc-Me in d6-DMSO.
The thermal stability of DUT-5Me and DUT-5Im was also investigated and compared to that of
DUT-5 (Fig. 3.9). The TGA plots indicate that DUT-5Me retained the thermal stability exhibited by DUT-
5, however DUT-5Im presents an extra region of ~10% mass loss between 230 and 310°C. In all cases,
the as synthesised samples lose 20-30% mass of DMF up to 210°C, while the activated samples lose
~5% mass of water within the same region. Water uptake occurs when transporting the activated
sample between instruments due the hygroscopic nature of these samples, which has been observed
60
to occur in several other systems,80 as well as in Chapter 2 here. Native DUT-5 degrades beyond ~450°C,
which is mirrored by both DUT-5Me and DUT-5Im.
Fig. 3.9: TGA for a) DUT-5Me as synthesised (dark green) and activated (light green) and b) DUT-5Im as synthesised (dark
blue) and activated (light blue), with comparison to DUT-5 as synthesised (red) on both plots. The grey dashed lines indicate
regions of interest as discussed in the main text above.
77 K N2 isotherms were collected for DUT-5Me and DUT-5Im in order to probe the effect of
the pendant functional groups on the available pore space within these DUT-5 analogues (Fig. 3.10).
Samples were activated as detailed in the experimental section (3.5.1) prior to collection. Data for
DUT-5 were extracted from the original work by Senkovska et al.15 using the WebPlotDigitizer
extension105 for Google Chrome, and replotted here. As expected, the inclusion of pendant
functionality within the pore space resulted in reduced N2 uptake. DUT-5 was reported with a
Brunauer-Emmett-Teller (BET) surface area of 1610 m2/g, and here DUT-5Me was calculated to have
BET surface area of 1550 m2/g, thus showing the minor drop in available surface area when a methyl
group is appended to the native bpdc linker. Furthermore, DUT-5Im has a BET surface area of only 870
m2/g, approximately half that of DUT-5.
61
Fig. 3.10: 77 K N2 isotherms for DUT-5 (red), DUT-5Me (green) and DUT-5Im (blue).
In order to provide greater understanding of the potential structure of DUT-5Me and DUT-
5Im, models were developed in the Accelrys program Materials Studio 5.0.81 The .cif provided by
Senkovska et al.15 was imported and used as a starting point. The PXRD data (Fig. 3.7 above) indicates
no change to the unit cell parameters, and thus the cell parameters were fixed in subsequent
optimisations. The appropriate pendant groups were manually added to bpdc at 100% occupancy to
form bpdc-Me or bpdc-Im linkers with the MOF. The modified structures were optimised by molecular
mechanics using the universal force field (UFF)106 to provide possible structures for DUT-5Me and DUT-
5Im (Fig. 3.11).
62
Fig. 3.11: Models of a) DUT-5, b) DUT-5Me, and c) DUT-5Im, showing the repeating Al chains joined by the appropriate linker
(left) and view down the pore windows of each MOF (right). Green, aluminium; blue, nitrogen, red, oxygen; grey, carbon.
Hydrogens are omitted for clarity.
The N2 adsorption isotherms (Fig. 3.10 above) had a large drop in uptake for DUT-5Im, and the
BET surface area calculated also fell by a factor of two from DUT-5 to DUT-5Im. This was consistent
with the calculated structure for DUT-5Im, which contained four imidazolium groups oriented into one
pore window (Fig. 3.11c). In order to maintain room within the MOF for metalation of the imidazolium
NHC precursor, or for gas adsorption, it appeared that the imidazolium functionality needed to be
diluted within the MOF. Thus a series of target ML-MOFs were proposed, whereby the bpdc-Im linker
would be included in stepwise loadings from 0 to 100%.
A one pot synthesis of a ML-DUT-5 analogue was recently reported by Krajnc et al.,97 where a
bipyridyl dicarboxylate (bpydc) linker was incorporated into native DUT-5. In that work, the bpydc and
bpdc were combined in a 1:5.5 ratio and the ML-DUT-5 synthesised following the conditions of
Senkovska et al.15 to achieve 12% incorporation of the bpydc linker (Fig. 3.12). Based upon this work,
a series of one pot reactions were attempted to develop a ML-MOF containing bpdc-Im. Despite
several attempts, no successful conditions could be found for the one-pot method. In a recent review
on ML-MOFs, Burrows93e noted that the solubility of linkers used must be similar for a successful one-
pot synthesis of ML-MOFs. Given this knowledge, an alternative approach was sought that does not
63
rely upon a one-pot approach. Solvent assisted linker exchange (SALE) was proposed to be more
favourable for synthesis of a ML-DUT-5 analogue, as both bpdc-Me and bpdc-Im have greater solubility
in DMF than bpdc.
Fig. 3.12: Representative image of bpydc incorporation into DUT-5 produced by Krajnc et al.97
Solvent assisted linker exchange allows control of the feed ratio of bpdc-Im, thus potentially
providing good control over the final amount of bpdc-Im incorporated into the new ML-DUT-5
analogue. The SALE process was initially attempted by addition of bpdc-Im to native DUT-5. In order
to follow bpdc-Im inclusion, the ML-MOFs were digested for NMR spectroscopy as detailed in the
experimental section (3.5.1). Study of the NMR spectra obtained became problematic as the aromatic
signals from bpdc were often obscured by the residual H3PO4 peak (Fig. 3.13). Both bpdc signals should
integrate to 4 H, however the signal closer to the acid peak integrates to 5 H prior to background
subtraction of the H3PO4 peak. While the residual acid peak could be background subtracted to yield
the correct integration for both bpdc signals, this is not reliably reproducible. Thus, measurement of
the ratio between integration of bpdc and bpdc-Im is not an accurate method for determining the
extent of bpdc-Im incorporation. An alternative method was to use DUT-5Me or DUT-5Im as the
starting point for SALE, thus allowing comparison of the diagnostic methyl peak for bpdc-Me with the
CH2 bridge signal in bpdc-Im. As both of these signals are well clear of the aromatic region (Fig. 3.8
above), this avoids the issues relating the residual H3PO4 peak.
64
Fig. 3.13: 1H NMR of DUT-5 in d6-DMSO with D3PO4 added. a) before background subtraction of the residual acid peak. This
causes the bpdc-2 signal to integrate incorrectly to 5H rather than 4.b) after background subtraction of the acid peak. The
bpdc-2 signal now integrates correctly to 4H, however this is not always reliably reproduced.
Renewed SALE efforts began with bpdc-Me added to pure DUT-5Im in an attempt to provide
ML-MOFs with high loading of bpdc-Im. Secondly, bpdc-Im was added to pure DUT-5Me to provide
ML-MOFs with low loadings of bpdc-Im. Unfortunately, the first approach did not result in any
inclusion of bpdc-Me measurable by NMR spectroscopy. The second approach, however, yielded two
ML-DUT-5 analogues. Five equivalents of bpdc-Im resulted in 25 ±5% inclusion of bpdc-Im, and two
equivalents yielded 5 ±2% inclusion (Table 3.2 and 3.3 for integration values, Appendix Fig. A2.3 to
A2.10). The overall SALE process is summarised in Fig. 3.14. The two new ML-DUT-5 analogues are
referred to herein as DUT-5Im25% and DUT-5Im5%, respectively.
Table 3.2: Linker incorporation details for SALE reactions from DUT-5Me with 5 equivalents of bpdc-Im. Loading calculated as detailed in the experimental section 3.5.4.
Sample number bpdc-Im methyl peak
integration
bpdc-Me methyl peak
integration bpdc-Im loading
1 1.34 3 22.3%
2 1.72 3 28.7%
3 1.40 3 23.3%
4 1.7 3 28.3%
65
Table 3.3: Linker incorporation for SALE reactions from DUT-5Me with 2 equivalents of bpdc-Im. Loading calculated as detailed in the experimental section 3.5.4.
Sample Number bpdc-Im methyl peak
integration measured
bpdc-Me methyl peak
integration bpdc-Im loading
1 0.42 3 7.0%
2 0.28 3 4.6%
3 0.27 3 4.5%
4 0.24 3 4.0%
Fig. 3.14: Summary of SALE reactions attempted for the DUT-5 systems.
The SALE reactions were intended to provide several ML-MOFs covering the range of 0-100 %
bpdc-Im occupancy. As the SALE reactions were only successful using DUT-5Me as the starting point,
and given that five equivalents only resulted 25% inclusion of bpdc-Im, higher loadings were not
pursued due to the large excess of bpdc-Im required. Despite the lack of quantitative substitution, it
has been shown in literature that low loadings of a modified linker is often desirable, especially for
catalytic applications, where overcrowding of the pore space would be an issue.42a
Synthesis of the ML-DUT-5 analogues was confirmed by both PXRD and TGA measurements
(Fig. 3.15). The PXRD data show a match of the key reflections for DUT-5Im to both DUT-5Im25% and
DUT-5Im5% and confirms that both ML-MOFs are isostructural to DUT-5Im, however the 25% loading
66
causes reduced peak intensity. All three samples were prepared for TGA as synthesised, and hence
exhibit DMF and water loss up to ~200°C. The extra mass loss observed for DUT-5Im compared to DUT-
5 (refer to Fig. 3.9 above) is observed for both ML-DUT-5 analogues, with the relative percentage mass
lost in this region comparable to the bpdc-Im loading calculated by NMR measurements. This region
of mass loss is postulated to be loss of the imidazolium group. The mass lost in this region corresponds
to 20%, 10%, and 5% of the final mass for DUT-5Im, DUT-5Im25%, and DUT-5Im5% respectively. Finally,
as for all the DUT-5 analogues presented in this work, the framework degrades above ~450°C.
Fig. 3.15: a) PXRD for DUT-5Im (blue), DUT-5Im25% (purple), and DUT-5Im5% (pink). Note reduced intensity was observed for
DUT-5Im25%. b) TGA for DUT-5Im as synthesised (blue), DUT-5Im25% (purple), and DUT-5Im5% (pink). The grey dashed lines
indicate regions of interest as discussed in the main text above.
3.3 Metalation of NHC-precursors
3.3.1 Molecular system
As discussed in Chapter 1, post synthetic metalation provides an opportunity to embed metal sites
within a MOF, for either catalysis or enhanced gas sorption and separation. Here the molecular system
was first studied to prove that bpdc-Im may be metalated, before pursuing metalation of the DUT-5
67
analogues. Using a facile synthesis recently reported for [(NHC)MX(cod)] complexes (cod =
cyclooctadiene, M = metal, X = halide) by Savka and Plenio,107 reaction of the bpdc-Im PF6 salt (3.4)
yielded an NHC-Rh molecular complex as detailed in the experimental (bpdc-Im@Rh). Single crystals
were isolated from a large amount of precipitate, and the final structure of bpdc-Im@Rh was revealed
by SCXRD (Fig. 3.16). As all material was used for crystallisation, and only a small quantity of single
crystals were obtained amongst the precipitate, no reaction yield or NMR data were collected. In the
interest of time this Rh metalation process was extended to DUT-5Im without further characterisation
of the molecular system. Any future work in this area should include complete characterization of the
molecular system.
Fig. 3.16: Structure of bpdc-Im@Rh. Note that the Rh is bonded to the cod group by its two double bonds, and the four Rh-C
single bonds are representative of this. C: grey, O: red, N: blue, Rh: yellow, Cl: green, H: white.
3.3.2 MOF system
Metalation with Rh of DUT-5Im, DUT-5Im25% and DUT-5Im5% was attempted following the same
conditions as for the molecular system. The MOFs were isolated from the sparingly soluble K2CO3 in
the reaction mixture by a density based separation with CHCl3, followed by washing of the MOFs with
acetone. EDX analysis revealed incorporation of Rh into the MOF as detailed in Table 3.4. Despite a
number of attempts, ICP-MS could not be performed reliably due to the resilience of the Al-based
frameworks to acidic digestion. The harshest conditions tested here were attempted dissolution in
aqua regia, which is a mixture with a 1:3 molar ratio of nitric acid to hydrochloric acid. Future work in
this area could entail use of hydrofluoric acid for acidic digest of the DUT-5Im materials, however this
will require highly specialised equipment and staff training which was not undertaken here.
68
Table 3.4: Extent of Rh metalation obtained for DUT-5Im (pure and ML-MOFs) as determined by EDX.
Sample Metalation Calculated by EDX (%)
DUT-5Im@Rh 30.2
DUT-5Im25%@Rh 24.8
DUT-5Im5%@Rh 11.1
Production of these MOFs was targeted toward Rh based catalysis, whereby the cod bound to the Rh
could be replaced by CO ligands as demonstrated by Savka and Plenio.107 Bubbling of CO gas through
the metalation reaction mixture in an attempt to replace the cod group resulted in destruction of all
three DUT-5Im analogues tested. This was observed to occur both before and after removal of the
K2CO3, and may be due to displacement of the bulky cod group disrupting the Al-carboxylate bonds of
the metal node, thus causing the framework to break down.
3.4 Conclusions Two new DUT-5 analogues, DUT-5Me and DUT-5Im have been synthesised from linkers bpdc-Me and
(33.6 mg, 0.24 mmol, 3 equiv.) were suspended in acetone (3 mL) and the reaction mixture stirred at
60°C for 16 h Solvent was removed in vacuo, the crude product was dissolved in toluene, filtered, and
73
toluene also removed in vacuo to yield bpdc-Im@Rh as a yellow solid. The compound was dissolved in
DCM (3 mL) and a layer of hexane (3 mL) added slowly above the DCM layer for crystallisation. The vial
was left in the dark for 2 days, then the cap loosened to allow slow removal of volatiles. Single yellow
crystals were formed over 18 h.
3.5.3 MOF Synthesis
DUT-5 was synthesised following the method detailed by Senkovska et al.15 DUT-5 analogues DUT-5Me
and DUT-5Im were synthesised with minor adjustments to the method for DUT-5 as follows:
DUT-5Me. N,N’-Dimethylformamide (5.0 mL) was added to a 20 mL Teflon capped glass vial containing
bpdc-Me (46.2 mg, 0.180 mmol, 1 equiv.) and Al(NO3)2·9H2O (87.8 mg, 0.234 mmol, 1.30 equiv.) and
the vial sonicated for 10 min to achieve full dissolution. The reaction mixture was heated at 120°C for
24 h, then allowed to cool to room temperature. The white precipitate was collected by centrifugation
(5000 rpm, 2 min.), washed five times with fresh DMF, and dried overnight in a desiccator to yield DUT-
5Me as a white powder (72.1 mg, 81%). νmax (neat, cm-1): 1615 (s, C=O), 1545 (s, C=C).
DUT-5Im. N,N’-Dimethylformamide (5.0 mL) was added to a 20 mL Teflon capped glass vial containing
bpdc-Im (60.5 mg, 0.145 mmol, 1 equiv.) and Al(NO3)2·9H2O (67.3 mg, 0.180 mmol, 1.24 equiv.) and
the vial sonicated for 10 min to achieve full dissolution. The reaction mixture was heated at 120°C for
24 h, then allowed to cool to room temperature. The white precipitate was collected by centrifugation
(5000 rpm, 2 min) and washed five times with fresh DMF and dried overnight in a desiccator to yield
DUT-5Im as a white powder (73.4 mg, 77%). νmax (neat, cm-1): 1608 (s, C=O), 1553 (s, C=C).
3.5.4 Solvent Assisted Linker Exchange
DUT-5Im25%. N,N’-Dimethylformamide (2.0 mL) was added to a 6 mL Teflon capped glass vial containing
DUT-5Me (0.020 g, 0.004 mmol, 1 equiv.) and bpdc-Im (0.084 g, 0.020 mmol, 5 equiv.) and the reaction
mixture stirred vigorously at 90°C for 18 h. Solvent was removed by centrifugation (5000 rpm, 2min.),
the solid washed five times with fresh DMF, and dried overnight in a desiccator to yield DUT-5Im25%.
The linker composition was measured by NMR spectroscopy averaged over four samples and
determined to consist of 75 ± 5 % bpdc-Me and 25 ± 5 % bpdc-Im.
74
DUT-5Im5%. N,N’-Dimethylformamide (2.0 mL) was added to a 6 mL Teflon capped glass vial containing
DUT-5Me (0.020 g, 0.004 mmol, 1 equiv.) and bpdc-Im (0.034 g, 0.008 mmol, 2 equiv.) and the reaction
mixture stirred vigorously at 90°C for 18 h. Solvent was removed by centrifugation (5000 rpm, 2min.)
and the remaining solid washed five times with fresh DMF, and dried overnight in a desiccator to yield
DUT-5Im5%. The ligand composition was measured by NMR spectroscopy averaged over four samples
and determined to be 95 ± 2 % compound bpdc-Me and 5 ± 2 % compound bpdc-Im.
Incorporation of bpdc-Im was determined by NMR spectroscopy in d6-DMSO as follows:
The integration of the bpdc-Im methyl peak at 3.72 ppm was measured, with the integration value
referenced to the 3H methyl peak of bpdc-Me at 2.26 ppm. At a theoretical bpdc-Im loading of 50%,
these two integration values will both have a value of 3H. Hence the integration value for the bpdc-Im
methyl peak was divided by 3 to determine how close to 50% loading of bpdc-Im was achieved. I.e. an
integration reading of 1.5H means that the sample is half way toward 50% loading of bpdc-Im.
Therefore to obtain the loading of bpdc-Im relative to the entire MOF system, this value was divided
by two. In the example given, an integration of 1.5H would therefore yield a final bpdc-Im loading of
25%, and this is shown in Table 3.5 below.
Table 3.5: Example calculations for NMR spectroscopy based determination of bpdc-Im incorporation into mixed DUT-5 systems.
Integration reading for bpdc-Im
peak at 3.72 ppm
Extent toward 50% loading of
bpdc-Im in the MOF
Occupancy within the entire
MOF system
2H 2/3 33.3%
1.5H 1/2 25%
1H 1/3 16.7%
75
Chapter 4: High Pressure Adsorption in DUT-5 Analogues
4.1 Introduction and Scope of Chapter
Imidazolium containing MOFs have been used in a range of gas adsorption studies, as discussed in the
Chapter 1. In Chapter 2 the presence of an imidazolium moiety within the pore space of Zn(II) and
Cu(II) based MOFs was shown to affect H2 adsorption enthalpy. Additional gas sorption studies on the
role of imidazolium moieties are now possible with the new materials DUT-5Me and DUT-5Im in hand,
particularly given that these frameworks have increased porosity compared to those described in
Chapter 2.
The original report of DUT-5 showed promise for gas storage and separation, specifically high
pressure sorption of H2, CO2, and CH4.15 More recently, the Kitagawa group demonstrated
incorporation of an ionic liquid (IL) into a MOF for the first time.63 Inclusion of ILs into MOFs can be
advantageous for gas storage and separation, particularly for CO2 or H2, as demonstrated both
computationally62a, 62b, 62d and experimentally.62c, 64b-d, 108 Other, less studied, potential applications
include catalysis, synthesis templates for nanoporous carbon-rich structures, or low temperature
batteries.64a, 64b
There are three main methods to incorporate ILs into MOFs that have been developed, as
summarised by Fujie and Kitagawa.64b The first relies on the target MOF possessing coordinatively
unsaturated metal sites, where the IL is added in solution and trapped by interaction with the vacant
coordination site.109 The second strategy is to synthesise the IL within the pores of the MOF, using the
MOF as a pseudo reaction vessel.110 This allows for incorporation of ILs that would not normally fit
through the pore openings, as only the starting materials need to be introduced. Thus the IL is trapped
within the pore where it is synthesised. The final strategy is to introduce the IL into the MOF by capillary
action, where the MOF is first activated, then stirred with the IL and heated to aid the diffusion
process.63-64, 108
There are no coordinatively unsaturated Al sites within DUT-5 or its analogues, and thus the
first method of IL incorporation is not available in these systems. The third method was chosen here,
as it provides a fast and relatively simple method for IL incorporation, and activation conditions for the
76
MOFs were already known from the work of Senkovska et al.15 It was thought that DUT-5Im may have
a more favourable interaction with ILs due to the charged moieties already present in the pore space
of DUT-5Im. Thus, ILs with imidazolium cations were chosen due to the presence of the imidazolium
functionality on the bpdc-Im linker. ILs with an imidazolium cation are commercially available with a
variety of counter ions including; thiocyanate, acetate, ethyl sulfate, chloride, or bromide. Often the
imidazolium cation is substituted at the N positions, and the identity of both the cation and anion in
an IL can significantly alter the melting point. An IL which is a liquid at room temperature removes the
need for a solvent when introducing the IL into a MOF by capillary action. The room temperature IL 1-
ethyl-3-methyl-imidazolium ethyl sulfate (EMIM-ES, Fig. 4.1) was chosen as it was relatively cheap and
expected to interact favourably with the imidazolium group present in DUT-5Im. Herein the
incorporation of EMIM-ES into DUT-5Me and DUT-5Im, and subsequent gas sorption measurements
are detailed.
Fig. 4.1: Chemical structure of EMIM-ES
4.2 Ionic Liquid Incorporation for High Pressure Adsorptions
Capillary action was used to incorporated EMIM-ES into DUT-5Im and DUT-5Me as detailed in Fujie’s
review64b and the experimental section to yield DUT-5Im@IL and DUT-5Me@IL respectively. The
samples were washed with ethanol while under reduced pressure to remove EMIM-ES on the MOF
exterior, leaving only EMIM-ES within the MOF pores. Inclusion of EMIM-ES was confirmed by FTIR and
1H NMR spectroscopy (FTIR spectra Fig. 4.2, 1H NMR Fig. 4.3 and Fig. 4.4, full FTIR spectra are available
in the Appendix, Fig. A3.1 and A3.2). FTIR spectra were collected for the MOFs both before and after
EMIM-ES was added, and the inclusion of EMIM-ES is shown by the additional peaks for EMIM-ES
present in the FTIR trace for DUT-5Me@IL and DUT-5Im@IL (Fig. 4.2). Integration of the 1H NMR
spectra revealed incorporation of 1.0 EMIM-ES molecules per linker for DUT-5Me, but only 0.4 EMIM-
ES molecules per linker in DUT-5Im. To determine EMIM-ES loading relative to available space, the N2
77
77K isotherm data collected in Chapter 3 was considered. This gave maximal uptake values (at ~710
mbar) of 508 and 292 cm3/g of N2 for DUT-5Me and DUT-5Im, respectively, indicating that DUT-5Im
has significantly less available space than DUT-5Me. Given the reduced space in DUT-5Im,
incorporation of EMIM-ES per linker was expected to be approximately half that seen for DUT-5Me.
Hence while the loading per linker is significantly different, the loading relative to available pore
volume is very similar.
Fig. 4.2: FTIR spectra for the DUT-5 analogues and EMIM-ES. The EMIM-ES trace (black) compared to a) DUT-5Me (green); b)
DUT-5Im (blue); c) DUT-5Me@IL (purple);and d) DUT-5Im@IL (orange) demonstrating the successful incorporation of EMIM-
ES. This can be seen by the EMIM-ES signals which are now reflected in the IL loaded MOFs (c and d) as compared to MOF only
(a and b). Note that DUT-5Im has some similarity to EMIM-ES due to the presence of imidazolium functionality in the bpdc-
Im linker.
78
Fig. 4.3: 1H NMR for a sample of DUT-5Me@IL. Samples were digested with D3PO4 in d6-DMSO, giving residual solvent peaks
at δ 8.24 and 2.50 ppm (D3PO4 and d6-DMSO respectively). Signals marked with a solid black star are due to EMIM-ES,
unmarked peaks are due to the bpdc-Me linker. The hollow star represents two EMIM-ES signals at δ 7.73 and 7.64 ppm, and
the signals are shown in the inset.
Fig. 4.4: 1H NMR for a sample of DUT-5Im@IL. Samples were digested with D3PO4 in d6-DMSO, giving residual solvent peaks
at δ 8.24 and 2.50 ppm (D3PO4 and d6-DMSO respectively). Signals marked with a solid black star are due to EMIM-ES,
unmarked peaks are due to the bpdc-Im linker. Note that the EMIM-ES signal at δ 3.71 ppm as seen in DUT-5Me@IL (Fig. 4.3)
is obscured by the 3H singlet due to the methyl group bound to the imidazolium in bpdc-Im. The hollow star represents two
EMIM-ES signals at δ 7.72 and 7.63 ppm, and the signals are shown in the inset.
79
4.3 High Pressure Measurements
As discussed earlier, IL incorporation into MOFs can be beneficial for gas storage and separation.
Senkovska et al.15 showed that DUT-5 has significant potential in high pressure adsorption of H2, CO2,
and CH4. Investigation of high pressure adsorption provides increased industrial relevance, as most
pre-combustion processes are performed at high pressure.55f Of particular note is the purification of
mixed CO2/CH4 gas streams in order to yield purified CH4 fuel. The low volumetric energy density of
CH4 means that high pressure is also required for storage applications in order to remain competitive
with regular gasoline fuels.55j
All modified DUT-5 frameworks (DUT-5Me, DUT-5Im, DUT-5Me@IL, and DUT-5Im@IL) were
investigated for high pressure sorption of CO2 and CH4, with representative uptake values and ratios
summarised in Tables 4.1-4.3 below. This analysis is the first time that high pressure gas adsorption
has been measured in any DUT-5 analogues, as well as the first time IL incorporation into DUT-5
analogues has been demonstrated.
DUT-5 exhibits a CO2 uptake of approximately 8 mmol/g at 10 bar according to Senkovska et
al.,15 which is approximately 180 cm3/g (Table 4.1 and Fig. 4.5a, see Appendix for calculations). This
was the maximal pressure recorded in the original report of DUT-5, and thus all comparisons to DUT-
5 made here are at a CO2 pressure of 10 bar. As in section 4.2, the N2 isotherms from Chapter 3 were
used to provide an estimate of the available volume within DUT-5Me and DUT-5Im, with uptake
volumes of 508 and 292 cm3/g N2 respectively. The N2 isotherm for DUT-5 exhibited a maximal uptake
of N2 of 555 cm3/g, suggesting that DUT-5Me may have a similar CO2 uptake to DUT-5 based solely on
available volume. However, DUT-5Me was measured to have CO2 uptake of 97 cm3/g at 10 bar, which
is approximately 50% of DUT-5’s CO2 uptake. This could indicate that the introduction of a methyl
group to the bpdc linker causes a significant reduction in affinity for CO2. Adsorption of CH4 uptake by
DUT-5Me results in a similar analysis to CO2 uptake, with DUT-5Me exhibiting only 40% of the CH4
uptake observed for DUT-5 at 35 bar (Fig. 4.5b).
80
To investigate the observed reduction in CO2 uptake for DUT-5Me further, the available pore
space of DUT-5, DUT-5Me and DUT-5Im was modelled in Materials Studio 5.0.81 These calculations
yielded pore volumes of 2090, 605, and 425 Å3, respectively (Table 4.1). These data suggest that DUT-
5Me would have CO2 uptake approximately one third that of DUT-5, which is slightly lower than that
observed experimentally. This may be due to assumptions made in modelling the structure of DUT-
5Me, where an even distribution of one methyl group per void space was assumed for simplicity.
However this may not occur experimentally, and a single crystal structure of DUT-5Me is required to
determine the exact methyl group positioning. Single crystals were not obtained here and have been
shown in literature to be very difficult to obtain for many other Al based MOFs.20b, 111
DUT-5Im exhibits a CO2 uptake of 84 cm3/g at 10 bar, which is just below 50% of the CO2 uptake
under equivalent conditions for DUT-5 (Table 4.1 and Fig. 4.6a). The 77K N2 maximal uptake for DUT-
5Im is 292 cm3/g, approximately half that of DUT-5. Meanwhile the calculated pore space for DUT-5Im
is 425 Å3, which is only 20% of the pore space available in DUT-5. For DUT-5Im the CO2 uptake predicted
by the N2 uptake volume was a good match to that observed experimentally. However the
experimental uptake was increased compared to that predicted based on the calculated pore space.
Similarly to DUT-5Me, this may be due to the assumption made in modelling the structure that one
imidazolium group from bpdc-Im lies within each void space of DUT-5Im, as opposed to the unknown
experimental distribution of imidazolium groups. The increased experimental uptake may also be due
to an increased CO2 affinity for DUT-5Im compared to DUT-5Me, however this cannot be confirmed
based on the data available and requires further investigation.
As discussed in Chapter 1 and section 4.1, ILs have been proposed for use in gas stream
purification due to the promising uptake of CO260 and SO2
61 that has been measured. Furthermore,
recent computational studies have demonstrated the potential of IL loaded MOFs for CO2 uptake.62a,
62b, 62d Experimentally, Ma et al.64c demonstrated that loading the MOF MIL-100-Cr with an IL resulted
in improved CO2/N2 selectivity. In this work, incorporation of EMIM-ES into both DUT-5Me and DUT-
5Im was intended to increase CO2 adsorption, however the reverse occurred (Table 4.2, Fig. 4.5c and
d, and Fig. 4.6c and d). DUT-5Me@IL exhibited only 40% of the DUT-5Me CO2 uptake at 35 bar, which
81
is most likely due to the DUT-5Me pore space being filled with EMIM-ES, leaving limited space to
adsorb CO2. The same reduction in uptake occurred for DUT-5Im, with DUT-5Im@IL adsorbing only
20% of the CO2 or CH4 adsorbed by DUT-5Im. Loading with EMIM-ES resulted in an improved CH4/CO2
selectivity from DUT-5Me to DUT-5Me@IL, but reduced the same selectivity from DUT-5Im to DUT-
5Im@IL (Table 3). While potentially interesting, the low volumetric uptakes obtained for both systems
should be noted as they may affect the selectivity observed and result in a spurious conclusion.
Table 4.1: Selected uptake values for all DUT-5 systems for both CO2 and CH4. Value marked with a star (*) was measured by Senkovska et al.15 Pore space values were calculated in Materials Studio, except for samples containing IL.
MOF Gas Uptake at 10 bar (cm3/g) Uptake at 35 bar (cm3/g) Pore space (Å3)
DUT-5 CO2 180* -
2090 CH4 165 180
DUT-5Me CO2 97 143
605 CH4 32 74
DUT-5Me@IL CO2 31 53
N/A CH4 8.6 32
DUT-5Im CO2 84 125
425 CH4 25 55
DUT-5Im@IL CO2 15 36
N/A CH4 5.4 11
Table 4.2: Uptake ratios for all DUT-5 systems at representative points for both CO2 and CH4.
Ratio or System CO2 at 10 bar CO2 at 35 bar CH4 at 10 bar CH4 at 35 bar
DUT-5Me/DUT-5 0.54 - 0.19 0.41
DUT-5Me@IL/DUT-5Me 0.32 0.37 0.27 0.43
DUT-5Im/DUT-5 0.47 - 0.15 0.31
DUT-5Im@IL/DUT-5Im 0.18 0.29 0.22 0.20
Table 4.3: Ratio of CH4 to CO2 uptake at 10 bar.
System CH4/CO2 uptake ratio at 10 bar
DUT-5Me 0.52
DUT-5Me@IL 0.60
DUT-5Im 0.44
DUT-5Im@IL 0.31
82
Fig. 4.6: High pressure adsorption isotherms of CO2 (circles) and CH4 (triangles) for DUT-5Im (blue) and DUT-5Im@IL (orange).
Filled and open symbols indicate adsorption and desorption respectively.
As discussed in section 4.2, the IL loadings achieved here were 1.0 and 0.4 EMIM-ES moieties
per linker for DUT-5Me@IL and DUT-5Im@IL, respectively. Incorporation of ILs to increase gas uptake
Fig. 4.5: High pressure adsorption isotherms of CO2 (circles) and CH4 (triangles) for DUT-5Me (green) and DUT-5Me@IL (purple). Filled and open symbols indicate adsorption and desorption respectively.
83
is reliant on a subsequent increase in charge density present within the MOF, which has been shown
to occur in other charged MOF systems.65f, 75 In this case, however, it appears that EMIM-ES has
completely blocked the MOF pores, and any benefit of the charge density is outweighed by the lack of
space available for gas molecule uptake. A similar phenomenon has been observed in the literature by
both Fujie et al.63 and Khan et al.112 In fact, Fujie et al. used the reduction in N2 uptake in a 77K N2
isotherm to quantify the amount of IL adsorbed into the MOF being studied (ZIF-8). Furthermore, Khan
et al. showed that a loading of greater than 50% IL resulted in a complete loss of porosity.
4.4 Conclusions EMIM-ES incorporation was measured as 1.0 and 0.4 moieties per linker for DUT-5Me@IL and DUT-
5Im@IL respectively, which is the maximum loading possible in this case. All four DUT-5 systems
discussed here adsorb CO2 preferentially over CH4 (Fig. 4.5c and d; and Fig. 4.6c and d), however the
inclusion of EMIM-ES reduces the CO2/CH4 selectivity and overall uptake of both CO2 and CH4. Further
study of ILs in the DUT-5 systems here must consider the loading of IL into MOF as a critical factor.
4.5 Experimental General methods, including FTIR and 1H NMR digestions, were performed as describe in Chapters 2
and 3 unless otherwise specified.
4.5.1 Ionic Liquid Incorporation into MOFs
DUT-5Me or DUT-5Im were activated at 160°C for 18 h to remove all DMF molecules, as confirmed by
1H NMR spectroscopy, and retention of crystallinity was confirmed by PXRD (Appendix, Fig. A3.3 to
A3.5). This was followed by static soaking of the MOF in EMIM-ES (50 mg MOF/mL EMIM-ES) for 18 h
at room temperature. The MOF samples were washed with copious amounts of ethanol under reduced
pressure on a 3 Å glass sinter to remove excess EMIM-ES, then dried in a desiccator under continuous
vacuum for 3 h to remove any ethanol from the exterior MOF surface. Complete removal of ethanol
was confirmed by 1H NMR spectroscopy, and incorporation of EMIM-ES was observed by FTIR and 1H
NMR spectroscopy as detailed in the main text (Fig. 4.3 and Fig. 4.4 in section 4.2).
84
4.5.2 Preparation for Gas Sorption Isotherms
Prior to high pressure adsorption measurements, the relevant MOF (120 mg, DUT-5Me or DUT-5Im)
was activated at 160°C under vacuum for 16 h. The IL loaded samples (120 mg, DUT-5Me@IL or DUT-
5Im@IL) were prepared as detailed in section 4.5.1, and dried at 120°C under vacuum for 6 h to remove
any water present prior to sorption measurements. The high pressure CO2 and CH4 experiments were
conducted at 298K up to 35 bar using a high pressure volumetric analyser (Micromeritics HPVA-100).
UHP grade (99.999%) CO2, CH4, and He were used for all measurements (He was used to purge air from
the system prior to all measurements).
85
Chapter 5: Conclusions and Future Directions
In Chapter 2, metalation with Cu(I) of the NHC precursor present in a Zn(II) MOF {[Zn2(μ2-
HCOO)(HL1)2](NO3)∙DMF}n (1)reported by Crees et al.76 was achieved concomitantly with synthesis of
1 to yield the material {[Zn2(μ2-HCOO)(HL1)1.6(L1-Cu-Br)0.4](NO3)0.6 0.75DMF}n (1-Cu). This metalation
was optimised to a maximal occupancy of 40% to yield a partially metalated MOF. The coordination
chemistry of H3L1 was shown to be highly versatile, with four new MOFs synthesised from Cu(II), Co(II),
and Mg(II) nitrate salts, and Mn(II) chloride to give {[Cu2(μ2-HCOO)(HL1)2](NO3)·1.75DMF}n (2) and
{[M3(HL1)2(H2O)2](NO3)∙DMF}n (where M = Co, 3; Mg,4; Mn, 5). MOFs 1-5 are 1-D, with only 1, 1-Cu,
and 2 showing permanent porosity. Subsequently 1, 1-Cu, and 2 were investigated for H2 adsorption,
with enthalpies of adsorption of -9.9, -9.1, and -9.7 kJmol-1, respectively. The accessibility of the linker
in MOFs 1-5 was limited due to the undulating 1-D chain structures observed, which pack together to
leave only minimal pore space. While 1, 1-Cu, and 2 proved to have potential application in H2
adsorption, the structure of the H3L1 linker was concluded to have a negative influence on the overall
MOF network. The imidazolium core causes H3L1 to possess a bent geometry and, combined with the
meta positioning of the carboxylic acids on the phenyl group, this causes the MOF networks
synthesised here to tend toward a 1-D structure.
In an attempt to obtain more porous 2-D or 3-D MOFs the coordination chemistry of H3L2 was
also studied. The carboxylic acid groups of H3L2 are now substituted on the phenyl rings at the para
position, thus potentially providing improved MOF porosity. The limited solubility of H3L2 in standard
MOF synthesis conditions caused a significant reduction in the versatility of H3L2 as compared to H3L1.
However, despite the lack of a new porous MOF after screening against a wide array of metal salts, a
closer investigation into the reaction of H3L2 and Zn(II) salts led to the synthesis of two MOFs, 6 and 7.
Compound 6 is similar to a Zn(II) MOF reported in the literature by Sen et al.,65e however inspection of
the crystallographic data suggested a metal node assignment of a disordered Zn4O node provides an
improved fit of the electron density map than the Zn8O node presented by Sen et al. MOF 7 is again
similar to 6, however an increased quantity of DMF used in the synthesis inhibits the formation of the
Zn4O node and instead disordered solvent is observed to bind to the mononuclear Zn centres. Further
86
characterisation of MOFs 6 and 7 is required in order to complete the investigation into Zn4O node
formation in these systems.
Finally, SALE in the highly porous material MnMOF1 was investigated as a method to
incorporate NHC precursors into a robust, porous, MOF. The linker H4L3 was chosen for this study as
it was predicted to be capable of replacing the native MnMOF1 linker, H2L4. While the linker moieties
in MnMOF1 have been shown to be readily accessible for metalation,46 and mixed linker analogues
have been synthesised from one pot reactions,22 efforts in this study did not yield any incorporation of
H4L3 into MnMOF1. This was likely due to the relative shape of H4L3 compared to H2L4, where rotation
about the methylene bridge causes the metal binding carboxylate groups of H4L3 to be in an
unfavourable position for replacement of H2L4.
Throughout Chapter 2, linkers were used which contained an imidazolium functionality in the
core of the linker. It was generally concluded that this imposes a linker geometry that is inconvenient
for linker inclusion into known MOFs, and furthermore generally inhibits the formation of 2-D or 3-D
porous MOFs. Chapters 3 and 4 focused on NHC precursor incorporation into MOFs with improved
stability and porosity over the MOFs synthesised in Chapter 2. Linkers with NHC precursors attached
as pendant groups are used in order to provide more linear linker geometries that are better suited to
substitution into known MOFs. As such, the majority of future work from Chapter 2 is included in
Chapters 3 and 4. However, an alternative method not studied here is a pillaring strategy to form highly
porous, NHC precursor containing MOFs. This strategy is exemplified by the work of Sen et al.,65h where
2-D, imidazolium containing, sheets were pillared with a series of bipyridine based linkers to yield
highly porous 3-D MOFs. Pursuit of this strategy has been demonstrated in other systems,113 and would
be an important future experiment here. The most promising targets for this strategy are MOFs 6 and
7, where coordinated DMF could be removed and replaced with a pillaring agents to yield highly porous
3-D MOFs. However, the crystallographic study of MOF 7 is incomplete due to unresolvable disorder
within the structure, and thus an improved dataset is desired. This would drive further investigation
into the exact nature of the inhibition of the Zn4O node formation observed, and allow for a more
complete pillaring study as described previously.
87
Chapter 3 focussed on incorporation of NHC precursors into the highly stable MOF DUT-5. Two
linkers were used, bpdc-Me and bpdc-Im, and DUT-5 analogues DUT-5Me and DUT-5Im synthesised
from these linkers. Bpdc-Im contains an imidazolium group attached to a biphenyl backbone as a
pendant group, thus allowing improved access to the NHC precursor within the pore space of DUT-
5Im. 77K N2 isotherms and modelling suggested that full loading of bpdc-Im may fill the entire pore
space of DUT-5Im, and thus dilution of the indazolium functionality was desired. This was achieved by
SALE reactions starting from DUT-5Me and adding bpdc-Im to yield two mixed linker DUT-5 analogues
with 5% and 25% loading of bpdc-Im, DUT5-Im5% and DUT-5Im25%, respectively. These SALE reactions
were successful due to the relative solubility in DMF of bpdc-Me and bpdc-Im compared to bpdc. Both
bpdc-Me and bpdc-Im are highly soluble in DMF, and thus are easily displaced from the MOF during
the SALE process. On the other hand, bpdc is poorly soluble in DMF and thus is more difficult to
exchange via SALE reactions.
The methyl ester precursor of bpdc-Im was used as a model system to demonstrate the
viability of NHC metalation in bpdc-Im. The methyl ester was metalated with Rh to yield bpdc-Im@Rh,
as confirmed by SCXRD. The same conditions were extended to DUT-5Im, DUT5-Im5%, and DUT-5Im25%
and appeared to be successful. Rh was observed to be incorporated by EDX measurements, however
in the absence of a crystal structure or additional structural characterisation (such as solid-state NMR
spectroscopy or EXAFS) the exact placement of the Rh centre within the MOF and its coordination
environment could not be confirmed. Any future work concerning Rh metalation should involve full
characterisation of the molecular system, and the nature of metalation of all three DUT-5 analogues
needs to be confirmed. This may be achieved by attempting dissolution for ICP-MS in hydrofluoric acid,
which was not available at the time this work was conducted. Measurement of the intense Kα emission
lines for Rh in EDX was not possible due to the limited beam power (up to 30 keV) on the SEM used.
As such, the weaker L emission lines were used, which results in a low signal to noise ratio and possible
errors in the metalation level determined by this technique. ICP-MS provides a quantitative
measurement of Al and Rh ions within a sample, whereas EDX yields only a qualitative ratio of Al to
88
Rh. Future work in this area could also investigate alternative metal centres to Rh in order to provide
different NHC-metal complexes which may be simpler to characterise than the system investigated
here.
An alternative target for catalysis is the NHC itself for use in organocatalysis within DUT-5Im,
DUT5-Im5%, and DUT-5Im25%. An example of NHC based organocatalysis within a MOF was recently
provided by Schumacher et al.,96 where the NHC was obtained by deprotonation of the NHC precursor
while attached to the Zr(IV) MOF UiO-67. The ML-DUT-5 analogues are likely to be the best targets for
organocatalysis as the pore space is less crowded than the pure DUT-5Im material. Within this project
tuning of the N-substituents on the NHC precursor will likely be necessary in order to stabilise the
carbene produced within the MOF.
Chapter 4 investigated high pressure adsorption of CO2 and CH4 in DUT-5 analogues DUT-5Me
and DUT-5Im. The uptake of both CO2 and CH4 was negatively affected by the inclusion of extra
functional groups within the pores as compared to DUT-5. The IL 1-ethyl-3-methyl-imidazolium ethyl
sulfate (EMIM-ES) was introduced into DUT-5Me and DUT-5Im in an attempt to improve CO2 uptake,
and hence improve CO2/CH4 selectivity. However the reverse occurred experimentally, with drastically
reduced uptake in both MOFs to the extent that the total gas adsorbed was negligible. Any future work
in this section must consider the IL loading within the MOF as a critical factor. In this case it appears
there was too much IL and the MOF pores were completely blocked, however a lower loading of IL
may alleviate this issue. One such method to incorporate less IL could be to use a solution of EMIM-ES
in a solvent than can be easily removed from the MOF (e.g. ethanol).
Further investigation is required into ILs in MOFs with a wider scope than the study presented
here. For example, the effect of using an IL that is a liquid at ambient room temperature (as in Chapter
4) against an IL that is solid at room temperature has not been thoroughly studied. Furthermore, the
exact nature of the IL within MOFs is not well characterized and needs further investigation. For
example, it is not clear if the IL remains in the liquid state within the MOF, and if this is dependent on
the number of IL molecules present within the MOF pores.
89
Overall future directions for MOFs containing NHCs and their precursors centre on application
for catalysis and gas sorption. The limited number of metalation studies in literature have shown
promising properties both in terms of enhanced catalysis and gas sorption, and thus there is scope for
further investigation. A promising metalation based project would be the inclusion of a NHC-Ru species
within a highly porous MOF, which could be synthesised to yield a MOF bound species that is similar
in design to Grubbs’ second generation catalyst. Furthermore, charge separation has been
demonstrated to be a critical factor in gas sorption, and further study of the interplay between azolium
and NHC-metal moieties within a MOF would be beneficial in this area.
90
Publications Produced During Candidature
1. Hydrogen adsorption in azolium and metalated N-heterocyclic carbene
containing MOFs
P. K. Capon; A. Burgun; C. J. Coghlan; R. S. Crees; C. J. Doonan; C. J. Sumby, CrystEngComm 2016, 18, 7003.
91
References
1. (a) S. S.-Y. Chui; S. M.-F. Lo; J. P. H. Charmant; A. G. Orpen; I. D. Williams, Science 1999, 283,
1148; (b) H. Li; M. Eddaoudi; M. O'Keeffe; O. M. Yaghi, Nature 1999, 402, 276; (c) S. Kitagawa; R.
Kitaura; S. Noro, Angew. Chem. Int. Ed. 2004, 43, 2334; (d) G. Ferey, Chem. Soc. Rev. 2008, 37, 191;
(e) H. Furukawa; K. E. Cordova; M. O'Keeffe; O. M. Yaghi, Science 2013, 341, 1230444; (f) A.
Schneemann; V. Bon; I. Schwedler; I. Senkovska; S. Kaskel; R. A. Fischer, Chem. Soc. Rev. 2014, 43,
6062; (g) M. Eddaoudi; D. F. Sava; J. F. Eubank; K. Adil; V. Guillerm, Chem. Soc. Rev. 2015, 44, 228; (h)
Y. Bai; Y. Dou; L. H. Xie; W. Rutledge; J. R. Li; H. C. Zhou, Chem. Soc. Rev. 2016, 45, 2327.
2. M. P. Attfield; P. Cubillas, Dalton Trans 2012, 41, 3869.
3. A. J. Howarth; Y. Liu; P. Li; Z. Li; T. C. Wang; J. T. Hupp; O. K. Farha, Nat. Rev. Mater. 2016, 1,
15018.
4. K. Sumida; J. Arnold, J. Chem. Educ. 2011, 88, 92.
5. (a) S. Kitagawa; M. Kondo, Bull. Chem. Soc. Jpn. 1998, 71, 1739; (b) S. Horike; S. Shimomura;
S. Kitagawa, Nat. Chem. 2009, 1, 695.
6. A. Burgun; R. S. Crees; M. L. Cole; C. J. Doonan; C. J. Sumby, Chem. Commun. 2014, 50,
11760.
7. (a) K. S. Park; Z. Ni; A. P. Cote; J. Y. Choi; R. Huang; F. J. Uribe-Romo; H. K. Chae; M. O'Keeffe;
O. M. Yaghi, Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 10186; (b) T. Devic; C. Serre, Chem. Soc. Rev.
2014, 43, 6097.
8. M. Eddaoudi; J. Kim; N. Rosi; D. Vodak; J. Wachter; M. O'Keeffe; O. M. Yaghi, Science 2002,
295, 469.
9. J. H. Cavka; S. Jakobsen; U. Olsbye; N. Guillou; C. Lamberti; S. Bordiga; K. P. Lillerud, J. Am.
Chem. Soc. 2008, 130, 13850.
10. J. L. C. Rowsell; O. M. Yaghi, J. Am. Chem. Soc. 2006, 128, 1304.
11. S. R. Batten; R. Robson, Angew. Chem. Int. Ed. 1998, 37, 1460.
12. H. K. Chae; D. Y. Siberio-Perez; J. Kim; Y. Go; M. Eddaoudi; A. J. Matzger; M. O'Keeffe; O. M.
Yaghi, Nature 2004, 427, 523.
92
13. H. Deng; S. Grunder; K. E. Cordova; C. Valente; H. Furukawa; M. Hmadeh; F. Gandara; A. C.
Whalley; Z. Liu; S. Asahina; H. Kazumori; M. O'Keeffe; O. Terasaki; J. F. Stoddart; O. M. Yaghi, Science
2012, 336, 1018.
14. (a) D. Rankine; A. Avellaneda; M. R. Hill; C. J. Doonan; C. J. Sumby, Chem. Commun. 2012, 48,
10328; (b) T. D. Keene; D. Rankine; J. D. Evans; P. D. Southon; C. J. Kepert; J. B. Aitken; C. J. Sumby; C.
J. Doonan, Dalton Trans 2013, 42, 7871; (c) D. Rankine; T. D. Keene; C. J. Sumby; C. J. Doonan,
CrystEngComm 2013, 15, 9722.
15. I. Senkovska; F. Hoffmann; M. Fröba; J. Getzschmann; W. Böhlmann; S. Kaskel, Microporous
Mesoporous Mater. 2009, 122, 93.
16. A. D. Burrows; C. G. Frost; M. F. Mahon; C. Richardson, Angew. Chem. Int. Ed. 2008, 47, 8482.
17. S. M. Cohen, Chem. Rev. 2012, 112, 970.
18. J. D. Evans; C. J. Sumby; C. J. Doonan, Chem. Soc. Rev. 2014, 43, 5933.
19. (a) L. Ma; J. M. Falkowski; C. Abney; W. Lin, Nat. Chem. 2010, 2, 838; (b) P. V. Dau; S. M.
Cohen, Chem. Commun. 2013, 49, 6128; (c) G. Q. Kong; X. Xu; C. Zou; C. D. Wu, Chem. Commun.
2011, 47, 11005.
20. (a) J. Liu; D. M. Strachan; P. K. Thallapally, Chem. Commun. 2014, 50, 466; (b) E. D. Bloch; D.
Britt; C. Lee; C. J. Doonan; F. J. Uribe-Romo; H. Furukawa; J. R. Long; O. M. Yaghi, J. Am. Chem. Soc.
2010, 132, 14382.
21. Y. Xu; N. A. Vermeulen; Y. Liu; J. T. Hupp; O. K. Farha, Eur. J. Inorg. Chem. 2016, 2016, 4345.
22. M. Huxley; C. J. Coghlan; A. Burgun; A. Tarzia; K. Sumida; C. J. Sumby; C. J. Doonan, Dalton
Trans 2016, 45, 4431.
23. L. Cavallo; C. S. J. Cazin, N-Heterocyclic Carbenes: An Introductory Overview. 2010; Vol. 32, p
1.
24. H. W. Wanzlick; H. J. Schonherr, Angew. Chem. Int. Ed. 1968, 7, 141.
25. W. A. Herrmann; M. Elison; J. Fischer; C. Kocher; G. R. J. Artus, Angew. Chem. Int. Ed. 1995,
34, 2371.
26. A. J. I. Arduengo; R. L. Harlow; M. Kline, J. Am. Chem. Soc. 1991, 113, 361.
93
27. J. R. Zoeller; V. H. Agreda; S. L. Cook; N. L. Lafferty; S. W. Polichnowski; D. M. Pond, Catal.
Today 1992, 13, 73.
28. G. J. Sunley; D. J. Watson, Catal. Today 2000, 58, 293.
29. (a) X. Wang; Z. Wang; K. Ding, Endeavors to Bridge the Gap between Homo- and
Heterogeneous Asymmetric Catalysis with Organometallics. In Bridging Heterogeneous And
Attempted Syntheses for UiO-67 Table A2.1: Summary of synthetic conditions attempted toward obtaining single crystals of UiO-67. All reactions were conducted at 120°C and for 24 h unless otherwise noted. All literature procedures were attempted with linkers bpdc, bpdc-Me, and bpdc-Im.
Sample Reference Linker
concentration
Linker to
ZrCl4 ratio DMF volume Modulator Extra Additive(s)
Result for this
author
UiO-67 original
synthesis
Cavka et al.1 bpdc; 9x10-3 M 1:1 25 ml None None No precipitate
UiO-67 SC Schaate et al.2 bpdc; 9x10-3 M 1:1 25 ml Benzoic acid
30 equiv.
None Amorphous
precipitate
UiO-67, KOH
soaked flask, loose
lid
Øien et al.3 bpdc; 5.0x10-2 M 1:1 20 ml Benzoic acid
30 equiv.
35% HCl 0.83 μl. 48
h reaction.
Amorphous
precipitate
UiO-67 with bpydc
linker
Gonzalez et al.4 bpydc 1.2x10-2 M 1:2 anhydrous
DMF; 80 ml
Benzoic acid
80 equiv.
Deionized water
128 μl. 120 h
reaction.
Amorphous
precipitate
UiO-67 powder Cohen et al.5 bpdc 3.5x10-2 M 1:1 3 ml Acetic acid 33
equiv.
None Successful
UiO-67Me direct
synthesis
bpdc-Me 4.1x10-2
M
1:1 3 ml None None Low
crystallinity
sample
UiO-67Im direct
synthesis
bpdc-Im 4.1x10-2
M
1:1 3 ml None None No precipitate
1. J. H. Cavka; S. Jakobsen; U. Olsbye; N. Guillou; C. Lamberti; S. Bordiga; K. P. Lillerud, J. Am. Chem. Soc. 2008, 130, 13850.
2. A. Schaate; P. Roy; A. Godt; J. Lippke; F. Waltz; M. Wiebcke; P. Behrens, Chemistry 2011, 17, 6643.
3. S. Øien; D. Wragg; H. Reinsch; S. Svelle; S. Bordiga; C. Lamberti; K. P. Lillerud, Cryst. Growth Des. 2014, 14, 5370.
4. M. I. Gonzalez; E. D. Bloch; J. A. Mason; S. J. Teat; J. R. Long, Inorg. Chem. 2015, 54, 2995.
5. H. Fei; S. M. Cohen, Chem. Commun. 2014, 50, 4810.
104
NMR Spectroscopy Data
As synthesised 1H NMR spectra for DUT-5Me and DUT-5Im
Fig. A2.1: 1H NMR digest for DUT-5Me as synthesised in d6-DMSO with D3PO4 (H3PO4 peak at ~6 ppm), showing successful
incorporation of bpdc-Me.
Fig. A2.2: 1H NMR digest for DUT-5Im as synthesised in d6-DMSO with D3PO4 (H3PO4 peak at ~6 ppm), showing successful
incorporation of bpdc-Im.
105
SALE reactions starting from DUT-5Me with five equivalents of bpdc-Im
Fig. A2.3: 1H NMR digest in d6-DMSO with D3PO4 for SALE entry 1 in Table 3.2 in Chapter 3.
Fig. A2.4: 1H NMR digest in d6-DMSO with D3PO4 for SALE entry 2 in Table 3.2 in Chapter 3.
Fig. A2.5: 1H NMR digest in d6-DMSO with D3PO4 for SALE entry 3 in Table 3.2 in Chapter 3.
106
Fig. A2.6: 1H NMR digest in d6-DMSO with D3PO4 for SALE entry 4 in Table 3.2 in Chapter 3.
SALE reactions starting from DUT-5Me with two equivalents of bpdc-Im
Fig. A2.7: 1H NMR digest in d6-DMSO with D3PO4 for SALE entry 1 in Table 3.3 in Chapter 3.
107
Fig. A2.8: 1H NMR digest in d6-DMSO with D3PO4 for SALE entry 2 in Table 3.3 in Chapter 3.
Fig. A2.9: 1H NMR digest in d6-DMSO with D3PO4 for SALE entry 3 in Table 3.3 in Chapter 3.
Fig. A2.10: 1H NMR digest in d6-DMSO with D3PO4 for SALE entry 4 in Table 3.3 in Chapter 3.
108
Appendix 3: Supplementary Material for Chapter 4
Full FTIR Spectra for DUT-5Me, DUT-5Im, DUT-5Me@IL, DUT-5Im@IL, and EMIM-ES
Fig A3.1: Full FTIR spectra for EMIM-ES (black), DUT-5Me (green), and DUT-5Im (blue). Note that EMIM-ES was collected in
a Nujol mull, seen by the large stretch just below 3000 cm-1.
Fig A3.2: Full FTIR spectra for EMIM-ES (black), DUT-5Me@IL (purple), and DUT-5Im@IL (orange). Note that EMIM-ES was
collected in a Nujol mull seen by the large stretch just below 3000 cm-1.
109
Application of ideal gas law to convert mmol/g to cm3/g
𝐼𝑑𝑒𝑎𝑙 𝑔𝑎𝑠 𝑙𝑎𝑤: 𝑃𝑉 = 𝑛𝑅𝑇
where P = pressure (atm), V = volume (L), N = no. of moles (mol) ,R = universal gas constant, 0.08206
𝐿.𝑎𝑡𝑚
𝑚𝑜𝑙.𝐾 and T = temperature (K)
𝑉 =𝑛𝑅𝑇
𝑃
𝐴𝑡 𝑠𝑡𝑎𝑛𝑑𝑎𝑟𝑑 𝑐𝑜𝑛𝑑𝑖𝑡𝑖𝑜𝑛𝑠: 𝑉 =𝑛 (0.08206
𝐿. 𝑎𝑡𝑚𝑚𝑜𝑙. 𝐾
) (273 𝐾)
1 𝑎𝑡𝑚
𝑉 ≈ 22.4 𝑛
𝐻𝑒𝑛𝑐𝑒, 𝑖𝑓 𝑛 = 8 𝑚𝑚𝑜𝑙
𝑉 ≈ 180 𝑐𝑚3
Activated 1H NMR spectra and PXRD patterns for DUT-5Me and DUT-5Im
Fig. A3.3: 1H NMR spectrum for an activated sample of DUT-5Me in d6-DMSO and D3PO4. Residual solvent peaks are
observed at 2.50 and 8.20 ppm, respectively. All other peaks are due to DUT-5Me, indicating full removal of DMF.
Fig. A3.4: 1H NMR spectrum for an activated sample of DUT-5Im in d6-DMSO and D3PO4. Residual solvent peaks are
observed at 2.50 and 8.20 ppm, respectively. All other peaks are due to DUT-5Im, indicating full removal of DMF.
110
Figure A3.5: PXRD for DUT-5 simulated (black), DUT-5Im activated (blue), and DUT-5Me activated (green), showing